基于遗传-变邻域搜索算法的堆垛机作业调度优化  

Optimisation of stacker job scheduling based on genetic-variable neighbourhood search algorithm

作　　者：叶胜挺 王凯[1] 曾中荣 杜俊文 张迪博 YE Shengting;WANG Kai;ZENG Zhongrong;DU Junwen;ZHANG Dibo(School of Mechatronic Engineering and Automation,Foshan University,Foshan 528000,China;Guangdong Strong Metal Technology Co.,Ltd.,Foshan 528000,China)

机构地区：[1]佛山科学技术学院机电工程与自动化学院,广东佛山528000 [2]广东世创金属科技股份有限公司,广东佛山528000

出　　处：《电子设计工程》2024年第19期6-10,15,共6页Electronic Design Engineering

基　　金：国家重点研发计划项目(2022YFB4702300);广东省高校现代陶瓷与铝型材装备重点实验室(2017KSYS012);佛山市核心技术攻关项目(1920001000409)。

摘　　要：基于缩短自动化立体仓库中堆垛机完成货物出入任务时间的目的,建立了堆垛机作业调度的作业时间模型,通过修补机制、校正机制和遗传-变邻域搜索算法的优化方法,使得在优化的过程中,确保出入库任务序列的长度相等,并能够复用同一批作业任务中因出库作业产生的空货位。12组作业任务的仿真实验结果表明,相较于变邻域搜索算法,遗传-变邻域搜索算法的收敛速度更快,其中3组作业任务解的质量更高,总耗时和相对耗时的优化率分别为11.57%~21.56%和35.85%~68.89%,且复用空货位可以缩短堆垛机的作业时间。Based on the purpose of shortening the time for the stacker to complete the cargo exit and entry task in the automated stereoscopic warehouse,the operation time model of the stacker job scheduling is established,through the optimization method of repair mechanism,correction mechanism and geneticvariable neighborhood search algorithm,during in the process of optimization,ensure the length of the task sequence in and out of the warehouse is equal,and the empty cargo space generated by the same batch of job tasks can be reused.The simulation results of 12 groups of tasks show that compared with the variable neighborhood search algorithm,the convergence speed of the genetic-variable neighborhood search algorithm is faster,the solution quality of the three groups of tasks is higher,the optimisation rates of 11.57%~21.56%and 35.85%~68.89%for total and relative time consuming respectively,the reuse of empty cargo space can reduce the operation time of the stacker.

关 键 词：自动化立体仓库 作业调度 校正机制 复用空货位 遗传-变邻域搜索算法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]